/[pcre]/code/trunk/doc/pcretest.txt
ViewVC logotype

Contents of /code/trunk/doc/pcretest.txt

Parent Directory Parent Directory | Revision Log Revision Log


Revision 71 - (show annotations) (download)
Sat Feb 24 21:40:24 2007 UTC (7 years, 5 months ago) by nigel
File MIME type: text/plain
File size: 16399 byte(s)
Load pcre-4.4 into code/trunk.

1 NAME
2 pcretest - a program for testing Perl-compatible regular
3 expressions.
4
5
6 SYNOPSIS
7 pcretest [-d] [-i] [-m] [-o osize] [-p] [-t] [source] [des-
8 tination]
9
10 pcretest was written as a test program for the PCRE regular
11 expression library itself, but it can also be used for
12 experimenting with regular expressions. This document
13 describes the features of the test program; for details of
14 the regular expressions themselves, see the pcrepattern
15 documentation. For details of PCRE and its options, see the
16 pcreapi documentation.
17
18
19 OPTIONS
20
21
22 -C Output the version number of the PCRE library, and
23 all available information about the optional
24 features that are included, and then exit.
25
26 -d Behave as if each regex had the /D modifier (see
27 below); the internal form is output after compila-
28 tion.
29
30 -i Behave as if each regex had the /I modifier;
31 information about the compiled pattern is given
32 after compilation.
33
34 -m Output the size of each compiled pattern after it
35 has been compiled. This is equivalent to adding /M
36 to each regular expression. For compatibility with
37 earlier versions of pcretest, -s is a synonym for
38 -m.
39
40 -o osize Set the number of elements in the output vector
41 that is used when calling PCRE to be osize. The
42 default value is 45, which is enough for 14 cap-
43 turing subexpressions. The vector size can be
44 changed for individual matching calls by including
45 \O in the data line (see below).
46
47 -p Behave as if each regex has /P modifier; the POSIX
48 wrapper API is used to call PCRE. None of the
49 other options has any effect when -p is set.
50
51 -t Run each compile, study, and match many times with
52 a timer, and output resulting time per compile or
53 match (in milliseconds). Do not set -t with -m,
54 because you will then get the size output 20000
55 times and the timing will be distorted.
56
57
58 DESCRIPTION
59
60 If pcretest is given two filename arguments, it reads from
61 the first and writes to the second. If it is given only one
62 filename argument, it reads from that file and writes to
63 stdout. Otherwise, it reads from stdin and writes to stdout,
64 and prompts for each line of input, using "re>" to prompt
65 for regular expressions, and "data>" to prompt for data
66 lines.
67
68 The program handles any number of sets of input on a single
69 input file. Each set starts with a regular expression, and
70 continues with any number of data lines to be matched
71 against the pattern.
72
73 Each line is matched separately and independently. If you
74 want to do multiple-line matches, you have to use the \n
75 escape sequence in a single line of input to encode the new-
76 line characters. The maximum length of data line is 30,000
77 characters.
78
79 An empty line signals the end of the data lines, at which
80 point a new regular expression is read. The regular expres-
81 sions are given enclosed in any non-alphameric delimiters
82 other than backslash, for example
83
84 /(a|bc)x+yz/
85
86 White space before the initial delimiter is ignored. A regu-
87 lar expression may be continued over several input lines, in
88 which case the newline characters are included within it. It
89 is possible to include the delimiter within the pattern by
90 escaping it, for example
91
92 /abc\/def/
93
94 If you do so, the escape and the delimiter form part of the
95 pattern, but since delimiters are always non-alphameric,
96 this does not affect its interpretation. If the terminating
97 delimiter is immediately followed by a backslash, for exam-
98 ple,
99
100 /abc/\
101
102 then a backslash is added to the end of the pattern. This is
103 done to provide a way of testing the error condition that
104 arises if a pattern finishes with a backslash, because
105
106 /abc\/
107
108 is interpreted as the first line of a pattern that starts
109 with "abc/", causing pcretest to read the next line as a
110 continuation of the regular expression.
111
112
113 PATTERN MODIFIERS
114
115 The pattern may be followed by i, m, s, or x to set the
116 PCRE_CASELESS, PCRE_MULTILINE, PCRE_DOTALL, or PCRE_EXTENDED
117 options, respectively. For example:
118
119 /caseless/i
120
121 These modifier letters have the same effect as they do in
122 Perl. There are others that set PCRE options that do not
123 correspond to anything in Perl: /A, /E, /N, /U, and /X set
124 PCRE_ANCHORED, PCRE_DOLLAR_ENDONLY, PCRE_NO_AUTO_CAPTURE,
125 PCRE_UNGREEDY, and PCRE_EXTRA respectively.
126
127 Searching for all possible matches within each subject
128 string can be requested by the /g or /G modifier. After
129 finding a match, PCRE is called again to search the
130 remainder of the subject string. The difference between /g
131 and /G is that the former uses the startoffset argument to
132 pcre_exec() to start searching at a new point within the
133 entire string (which is in effect what Perl does), whereas
134 the latter passes over a shortened substring. This makes a
135 difference to the matching process if the pattern begins
136 with a lookbehind assertion (including \b or \B).
137
138 If any call to pcre_exec() in a /g or /G sequence matches an
139 empty string, the next call is done with the PCRE_NOTEMPTY
140 and PCRE_ANCHORED flags set in order to search for another,
141 non-empty, match at the same point. If this second match
142 fails, the start offset is advanced by one, and the normal
143 match is retried. This imitates the way Perl handles such
144 cases when using the /g modifier or the split() function.
145
146 There are a number of other modifiers for controlling the
147 way pcretest operates.
148
149 The /+ modifier requests that as well as outputting the sub-
150 string that matched the entire pattern, pcretest should in
151 addition output the remainder of the subject string. This is
152 useful for tests where the subject contains multiple copies
153 of the same substring.
154
155 The /L modifier must be followed directly by the name of a
156 locale, for example,
157
158 /pattern/Lfr
159
160 For this reason, it must be the last modifier letter. The
161 given locale is set, pcre_maketables() is called to build a
162 set of character tables for the locale, and this is then
163 passed to pcre_compile() when compiling the regular expres-
164 sion. Without an /L modifier, NULL is passed as the tables
165 pointer; that is, /L applies only to the expression on which
166 it appears.
167
168 The /I modifier requests that pcretest output information
169 about the compiled expression (whether it is anchored, has a
170 fixed first character, and so on). It does this by calling
171 pcre_fullinfo() after compiling an expression, and output-
172 ting the information it gets back. If the pattern is stu-
173 died, the results of that are also output.
174
175 The /D modifier is a PCRE debugging feature, which also
176 assumes /I. It causes the internal form of compiled regular
177 expressions to be output after compilation. If the pattern
178 was studied, the information returned is also output.
179
180 The /S modifier causes pcre_study() to be called after the
181 expression has been compiled, and the results used when the
182 expression is matched.
183
184 The /M modifier causes the size of memory block used to hold
185 the compiled pattern to be output.
186
187 The /P modifier causes pcretest to call PCRE via the POSIX
188 wrapper API rather than its native API. When this is done,
189 all other modifiers except /i, /m, and /+ are ignored.
190 REG_ICASE is set if /i is present, and REG_NEWLINE is set if
191 /m is present. The wrapper functions force
192 PCRE_DOLLAR_ENDONLY always, and PCRE_DOTALL unless
193 REG_NEWLINE is set.
194
195 The /8 modifier causes pcretest to call PCRE with the
196 PCRE_UTF8 option set. This turns on support for UTF-8 char-
197 acter handling in PCRE, provided that it was compiled with
198 this support enabled. This modifier also causes any non-
199 printing characters in output strings to be printed using
200 the \x{hh...} notation if they are valid UTF-8 sequences.
201
202 If the /? modifier is used with /8, it causes pcretest to
203 call pcre_compile() with the PCRE_NO_UTF8_CHECK option, to
204 suppress the checking of the string for UTF-8 validity.
205
206
207 CALLOUTS
208
209 If the pattern contains any callout requests, pcretest's
210 callout function will be called. By default, it displays the
211 callout number, and the start and current positions in the
212 text at the callout time. For example, the output
213
214 --->pqrabcdef
215 0 ^ ^
216
217 indicates that callout number 0 occurred for a match attempt
218 starting at the fourth character of the subject string, when
219 the pointer was at the seventh character. The callout func-
220 tion returns zero (carry on matching) by default.
221
222 Inserting callouts may be helpful when using pcretest to
223 check complicated regular expressions. For further informa-
224 tion about callouts, see the pcrecallout documentation.
225
226 For testing the PCRE library, additional control of callout
227 behaviour is available via escape sequences in the data, as
228 described in the following section. In particular, it is
229 possible to pass in a number as callout data (the default is
230 zero). If the callout function receives a non-zero number,
231 it returns that value instead of zero.
232
233
234 DATA LINES
235
236 Before each data line is passed to pcre_exec(), leading and
237 trailing whitespace is removed, and it is then scanned for \
238 escapes. Some of these are pretty esoteric features,
239 intended for checking out some of the more complicated
240 features of PCRE. If you are just testing "ordinary" regular
241 expressions, you probably don't need any of these. The fol-
242 lowing escapes are recognized:
243
244 \a alarm (= BEL)
245 \b backspace
246 \e escape
247 \f formfeed
248 \n newline
249 \r carriage return
250 \t tab
251 \v vertical tab
252 \nnn octal character (up to 3 octal digits)
253 \xhh hexadecimal character (up to 2 hex digits)
254 \x{hh...} hexadecimal character, any number of digits
255 in UTF-8 mode
256 \A pass the PCRE_ANCHORED option to pcre_exec()
257 \B pass the PCRE_NOTBOL option to pcre_exec()
258 \Cdd call pcre_copy_substring() for substring dd
259 after a successful match (any decimal number
260 less than 32)
261 \Cname call pcre_copy_named_substring() for substring
262
263 "name" after a successful match (name termin-
264 ated by next non alphanumeric character)
265 \C+ show the current captured substrings at callout
266 time
267 \C- do not supply a callout function
268 \C!n return 1 instead of 0 when callout number n is
269 reached
270 \C!n!m return 1 instead of 0 when callout number n is
271 reached for the nth time
272 \C*n pass the number n (may be negative) as callout
273 data
274 \Gdd call pcre_get_substring() for substring dd
275 after a successful match (any decimal number
276 less than 32)
277 \Gname call pcre_get_named_substring() for substring
278 "name" after a successful match (name termin-
279 ated by next non-alphanumeric character)
280 \L call pcre_get_substringlist() after a
281 successful match
282 \M discover the minimum MATCH_LIMIT setting
283 \N pass the PCRE_NOTEMPTY option to pcre_exec()
284 \Odd set the size of the output vector passed to
285 pcre_exec() to dd (any number of decimal
286 digits)
287 \Z pass the PCRE_NOTEOL option to pcre_exec()
288 \? pass the PCRE_NO_UTF8_CHECK option to
289 pcre_exec()
290
291 If \M is present, pcretest calls pcre_exec() several times,
292 with different values in the match_limit field of the
293 pcre_extra data structure, until it finds the minimum number
294 that is needed for pcre_exec() to complete. This number is a
295 measure of the amount of recursion and backtracking that
296 takes place, and checking it out can be instructive. For
297 most simple matches, the number is quite small, but for pat-
298 terns with very large numbers of matching possibilities, it
299 can become large very quickly with increasing length of sub-
300 ject string.
301
302 When \O is used, it may be higher or lower than the size set
303 by the -O option (or defaulted to 45); \O applies only to
304 the call of pcre_exec() for the line in which it appears.
305
306 A backslash followed by anything else just escapes the any-
307 thing else. If the very last character is a backslash, it is
308 ignored. This gives a way of passing an empty line as data,
309 since a real empty line terminates the data input.
310
311 If /P was present on the regex, causing the POSIX wrapper
312 API to be used, only B, and Z have any effect, causing
313 REG_NOTBOL and REG_NOTEOL to be passed to regexec() respec-
314 tively.
315 The use of \x{hh...} to represent UTF-8 characters is not
316 dependent on the use of the /8 modifier on the pattern. It
317 is recognized always. There may be any number of hexadecimal
318 digits inside the braces. The result is from one to six
319 bytes, encoded according to the UTF-8 rules.
320
321
322 OUTPUT FROM PCRETEST
323
324 When a match succeeds, pcretest outputs the list of captured
325 substrings that pcre_exec() returns, starting with number 0
326 for the string that matched the whole pattern. Here is an
327 example of an interactive pcretest run.
328
329 $ pcretest
330 PCRE version 4.00 08-Jan-2003
331
332 re> /^abc(\d+)/
333 data> abc123
334 0: abc123
335 1: 123
336 data> xyz
337 No match
338
339 If the strings contain any non-printing characters, they are
340 output as \0x escapes, or as \x{...} escapes if the /8
341 modifier was present on the pattern. If the pattern has the
342 /+ modifier, then the output for substring 0 is followed by
343 the the rest of the subject string, identified by "0+" like
344 this:
345
346 re> /cat/+
347 data> cataract
348 0: cat
349 0+ aract
350
351 If the pattern has the /g or /G modifier, the results of
352 successive matching attempts are output in sequence, like
353 this:
354
355 re> /\Bi(\w\w)/g
356 data> Mississippi
357 0: iss
358 1: ss
359 0: iss
360 1: ss
361 0: ipp
362 1: pp
363
364 "No match" is output only if the first match attempt fails.
365
366 If any of the sequences \C, \G, or \L are present in a data
367 line that is successfully matched, the substrings extracted
368 by the convenience functions are output with C, G, or L
369 after the string number instead of a colon. This is in addi-
370 tion to the normal full list. The string length (that is,
371 the return from the extraction function) is given in
372 parentheses after each string for \C and \G.
373
374 Note that while patterns can be continued over several lines
375 (a plain ">" prompt is used for continuations), data lines
376 may not. However newlines can be included in data by means
377 of the \n escape.
378
379
380 AUTHOR
381
382 Philip Hazel <ph10@cam.ac.uk>
383 University Computing Service,
384 Cambridge CB2 3QG, England.
385
386 Last updated: 20 August 2003
387 Copyright (c) 1997-2003 University of Cambridge.

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12